One Provider, Forty Programs, One Completion Rate. What That Average Hides

A large provider publishes one completion rate, one placement rate, and one median earnings figure. Behind those three numbers sit dozens of programs across several campuses, some running for eight months and some for three years, some with waiting lists and some the admissions team is quietly trying to fill. The average is real. It is also nearly useless for deciding whether to sign up for the one program you care about.

That gap is not an accident, and it is not concealment. It is the shape the reporting rules were built to produce.

The numbers were designed to answer a regulator, not an applicant

Institutional disclosure grew up in pieces, each piece added after something went wrong. Campus crime statistics arrived after a specific failure. Graduation rate reporting arrived because federal aid was flowing into schools where large numbers of students left with debt and no credential. Loan default reporting arrived to catch schools whose students could not repay. Earnings and completion disclosures for career programs arrived because job-placement claims in advertising had gotten loose.

Each rule was written to answer a question a regulator had, at the level a regulator could enforce. The Department of Education oversees the aid programs those disclosures attach to, and its unit of analysis is generally the institution, because that is the unit that receives funding and can lose it.

So the metrics report at the institution level, on a cohort clock, in a format that allows one school to be compared with another. None of those design choices serve someone deciding between the surgical technology program and the HVAC program at the same provider. They were never meant to.

What a large provider's average is actually averaging

The bigger the provider, the more the published rate is a weighted blend of things that have nothing to do with each other.

Consider what typically sits inside a single institutional completion figure at a multi-campus provider:

• Programs of very different lengths. A short certificate and a two-year associate degree finish on different clocks and are held to different completion windows.
• Programs with entrance screens and programs without. A nursing track with a competitive selection process will complete at a different rate than an open-enrollment business certificate, because the selection already happened.
• Campuses with different labor markets. The same welding curriculum in a region with a hiring backlog behaves differently than in one without.
• Online and in-person sections of the same course, often blended into one number.
• New programs launched last year, carrying almost no cohort history but folded into the average anyway.

The result is a figure that moves when a large program grows or shrinks, independent of whether any individual program got better or worse. A provider can improve every single program and watch its headline number fall because enrollment shifted toward a longer, harder one. Good providers know this and will say so if you ask.

The cohort rules decide who is allowed to count

This is where most of the surprise lives. Federal graduation rate reporting is built around a cohort of first-time, full-time students tracked over a defined period. That definition does a lot of work.

Students who transfer in with prior credit are commonly outside the counted cohort. So are students who start part-time, which at a career-focused provider can be a large share of the population: people keeping a job while they train. Students who leave for military service or certain other reasons may be removed from the denominator rather than counted as non-completers. A student who finishes the credential a semester after the measurement window closes is counted as a failure to complete.

None of that is fraud. It is the definition. But it means two honest facts can coexist:

The published completion rate covers a minority of the students actually enrolled, and the provider's own internal completion tracking, which counts everybody, shows a different number.

Larger providers usually have that internal number, because they need it to manage instructor load and campus capacity. Ask for it. Ask which students it includes that the federal figure excludes. A provider with a functioning institutional research office will answer in specifics.

Placement and earnings answer narrower questions than they appear to

"Job placement rate" is not one standardized measure. Depending on who requires it, the state licensing board, the programmatic accreditor, or the school itself, the rate can differ on every axis that matters:

• Who counts as placed. Any employment, employment in a related field, or employment requiring the credential. These produce very different numbers from the same graduating class.
• When it is measured. At 90 days, at six months, at one year. Short windows favor fields that hire fast.
• Who is in the denominator. Graduates only, or all starters. Available-for-work exclusions can remove students who moved, continued their education, or did not respond to the survey.
• Whether self-employment and part-time work count. In trades and cosmetology, this can shift a rate substantially.

Earnings figures carry their own boundaries. Aggregate earnings data tied to program completers reflects a specific cohort, in specific years, in whatever regions those graduates ended up. It is a rearview measurement of people who enrolled several years before you will. For a fast-moving field, that is a description of a labor market that has already changed.

What to ask for before you sign an enrollment agreement

Every question below has an answer that exists in the provider's records. A large organization is more likely to have them cleanly documented, not less.

1. The completion, placement, and licensure pass rates for this program, at this campus, for the last three cohorts. Not the institution figure.
2. The definition attached to each of those rates, in writing, including the measurement window and the exclusions.
3. The size of the cohort. A 90 percent rate on eleven students is a different claim than 90 percent on four hundred.
4. Which employers actually hired last year's graduates, and how many each took.
5. Whether the credential is accepted by the state board or certifying body you intend to sit with, confirmed against that body's own list rather than the brochure.
6. What happens to your tuition and credits if the program is discontinued or the campus consolidates.

The published numbers are a floor of verified information, standardized enough to be compared and audited. Treat them as the starting point of the conversation, and the specific answers as the thing you are actually buying with your decision.